How-To
Practical, no-fluff guides on AI agents, data engineering, and analytics — written from real client work, not theory.
How to Create Dashboards That Generate Revenue
Tim breaks down "2nd order modeling" — a simple shift in how you design a dashboard that turns it from something people glance at once into something that actually drives revenue decisions.
How to Rally Budget for an Analytics Project
A practical walkthrough of LTV, CAC, RFM, and Buy Till You Die modeling — and how to use them to build the internal case for funding an analytics project.
How to Build an Agency Strategist with ChatGPT and LangChain
An early, hands-on look at agentic AI — can ChatGPT and LangChain be combined into a working "agency strategist" that actually holds up to real agency work?
How to Write AI Agent Prompts That Actually Work
Most agent prompts fail for the same three reasons. Here is how we write agent personalities, engineer context, use ReAct loops, run agents in parallel worktrees, and hand off between sessions with a CLAUDE.md file, with real examples you can copy.
How to Build a Marketing Data Warehouse in 30 Days
A step-by-step look at how Latticework stands companies up on a modern marketing data warehouse — Fivetran, Snowflake, and a real dashboard — in under 30 days.
How to Get Started With Harness Engineering
The engineering discipline behind reliable AI agents: evals, observability, and the scaffolding that keeps them production-ready.
How to Set Up a Statistically Significant A/B Test
The sample-size math, guardrails, and common pitfalls that separate a real result from noise.
How to Do LTV:CAC Modeling Using Snowflake
Building a defensible LTV:CAC model directly in your warehouse — no spreadsheets, no guesswork.
How to Run a Marketing Mix Model Without a PhD in Statistics
A practitioner's guide to MMM — what it actually measures, and how to act on it.
How to Create Cohorts and Retention Curves with Claude
The retention math every retail brand should track, and how to turn cohort curves into a growth lever.
Ready when you are
Talk to us directly, or answer 3 quick questions and we’ll tell you exactly where to start.